The present invention relates to tricarbonyl compounds, and, in particular, to xcex1,xcex2-diketo acids, esters, and amides and a method of synthesis thereof. In addition, the present invention also relates to a method for the modular development of tricarbonyl compounds that have selected properties for a particular application. The xcex1,xcex2-diketo compounds are developed from (cyanomethylene)phosphoranes, carboxylic acids, and nucleophiles that include selected substituent groups that provide the desired properties in the xcex1,xcex2-diketo compounds. The iterative application of the method of the invention facilitates the synthesis of compounds having selected properties to meet the requirements of the particular application.
The discovery of new molecules has traditionally focused in two broad areas, biologically active chemical compounds, which are used as drugs for the treatment of life-threatening diseases, and new materials, which are used in commercial, and, especially, in high technological applications. In both areas, the strategy used to discover new compounds has involved two basic operations: the more or less random choice of a molecular candidate, prepared either via chemical synthesis or isolated from natural sources, and the testing of the molecular candidate for useful properties. This discovery cycle is repeated indefinitely until a molecule of a compound possessing the desirable property, i.e., a xe2x80x9clead moleculexe2x80x9d, is isolated or synthesized. This xe2x80x9clead moleculexe2x80x9d discovery process is inherently ad hoc in nature, and is time-consuming, laborious, unpredictable and costly.
Once a candidate lead molecule has been determined for a particular application, the synthetic chemist must subsequently find ways to synthesize structural variants of the lead molecule to optimize its properties for the application. In the case where the lead molecule is a synthesized organic species or a natural product, the chemist is usually limited to certain structural and synthetic reaction schemes. These schemes are dictated largely by the structural composition of the lead molecule and by the specific requirements of the application. For example, in cases where the lead molecule possesses a functionally important aromatic ring, various electrophilic and nucleophilic substitutions may be carried out on the ring to produce variants. However, each such case must be approached as a specific independent design and synthesis problem, starting each time from the beginning, because of the lack of availability of an appropriate chemistry to simply alter the structure of the lead compound to produce the variant.
Recently, some attempts have been made to modularize certain synthetic organic reaction schemes to facilitate modification and transformation of a lead or base compound. See, e.g., 1993 Proc. Natl. Acad. Sci. USA, 90, 6909. However, the molecules that can be produced by such attempts are extremely limited in their achievable diversity, and are still bounded by factors dictated by the choice of specific structural themes. In the case where the lead molecule is a naturally occurring, biological molecule, such as a peptide, a protein, an oligonucleotide or a carbohydrate, simple synthetic point-modifications to the lead molecule to produce variants are quite difficult to achieve.
A brief account of the strategies and tactics used in the discovery of new molecules is described below. Although the emphasis of the discussion is on molecules of biological interest, the technical problems encountered in the discovery of biologically active molecules is also illustrative of the problems encountered in the discovery of molecules that can serve as building blocks for the development of new tools and materials for a variety of high technological applications. Furthermore, as discussed below, these problems are also illustrative of the problems encountered in the development of fabricated structures and materials for high technological applications.
Modern theories of biological activity state that biological activities and, therefore, physiological states are the result of molecular recognition events. For example, nucleotides can form complementary base pairs so that complementary single-stranded molecules hybridize, resulting in double- or triple-helical structures that appear to be involved in regulation of gene expression. In another example, a biologically active molecule, referred to as a ligand, binds with another molecule, usually a macromolecule referred to as ligand-acceptor (e.g., a receptor, an enzyme, etc.), and this binding elicits a chain of molecular events which ultimately gives rise to a physiological state, e.g., normal cell growth and differentiation, abnormal cell growth leading to carcinogenesis, blood-pressure regulation, nerve-impulse generation and propagation, etc. The binding between ligand and ligand-acceptor is geometrically characteristic and extraordinarily specific, involving appropriate three-dimensional structural arrangements and chemical interactions.
A currently favored strategy for the development of agents which can be used to treat diseases involves the discovery of forms of ligands of biological receptors, enzymes, or related macromolecules, which mimic such ligands and either boost, i.e., agonize, or suppress, i.e., antagonize, the activity of the ligand. The discovery of such desirable ligand forms has traditionally been carried out either by random screening of molecules (produced through chemical synthesis or isolated from natural sources), or by using a so-called xe2x80x9crationalxe2x80x9d approach involving identification of a lead-structure, usually the structure of the native ligand, and optimization of its properties through numerous cycles of structural redesign and biological testing. Since most useful drugs have been discovered not through the xe2x80x9crationalxe2x80x9d approach, but through the screening of randomly chosen compounds, a hybrid approach to drug discovery has recently emerged which is based on the use of combinatorial chemistry to construct huge libraries of randomly-built chemical structures which are screened for specific biological activities. Brenner et al., 1992, Proc. Natl. Acad. Sci. USA, 89, 5381.
Most lead-structures which have been used in the xe2x80x9crationalxe2x80x9d drug design approach are native polypeptide ligands of receptors or enzymes. The majority of polypeptide ligands, especially the small ones, are relatively unstable in physiological fluids, due to the tendency of the peptide bond to undergo facile hydrolysis in acidic media or in the presence of peptidases. Thus, such ligands are decisively inferior in a pharmacokinetic sense to non-peptidic compounds, and are not favored as drugs. An additional limitation of small peptides as drugs is their low affinity for ligand acceptors. This phenomenon is in sharp contrast to the affinity demonstrated by large, folded polypeptides, e.g., proteins, for specific acceptors, such as receptors or enzymes, which often exist in the sub-nanomolar concentration range. For peptides to become effective drugs, they must be transformed into non-peptidic organic structures, i.e., peptide mimetics, which bind tightly, preferably in the nanomolar range, and can withstand the chemical and biochemical rigors of coexistence with biological tissues and fluids.
Despite numerous incremental advances in the art of peptidomimetic design, no general solution to the problem of converting a polypeptide-ligand structure to a peptidomimetic has been defined. At present, xe2x80x9crationalxe2x80x9d peptidomimetic design is done on an ad hoc basis. Using numerous redesign/synthesis/screening cycles, peptidic ligands belonging to a certain biochemical class have been converted by groups of organic chemists and pharmacologists to specific peptidomimetics. However, in the majority of cases, results in one biochemical area, such as, peptidase inhibitor design using the enzyme substrate as a lead, cannot be transferred for use in another area, such as, tyrosine-kinase inhibitor design using the kinase substrate as a lead.
In many cases, the peptidomimetics that result from a peptide structural lead using the xe2x80x9crationalxe2x80x9d approach comprise unnatural alpha-amino acids. Many of these mimetics exhibit several of the troublesome features of native peptides, which also comprise alpha-amino acids, and are, thus, not favored for use as drugs. Recently, fundamental research on the use of non-peptidic scaffolds, such as steroidal or sugar structures, to anchor specific receptor-binding groups in fixed geometric relationships have been described. See, e.g., Hirschmann et al., 1992 J. Am. Chem. Soc., 114, 9699-9701, and Hirschmann et al., 1992 J. Am. Chem. Soc., 114, 9217-9218. However, the success of this approach remains to be seen.
In an attempt to accelerate the identification of lead-structures, and also the identification of useful drug candidates through screening of randomly chosen compounds, researchers have developed automated methods for the generation of large combinatorial libraries of peptides and certain types of peptide mimetics, e.g., xe2x80x9cpeptoidsxe2x80x9d, which are screened for a desirable biological activity. For example, the method of Geysen, 1984 Proc. Natl. Acad. Sci. USA, 81, 3998, employs a modification of the Merrifield peptide synthesis, wherein the C-terminal amino acid residues of the peptides to be synthesized are linked to solid-support particles shaped as polyethylene pins. These pins are treated individually or collectively in sequence to introduce additional amino-acid residues forming the desired peptides. The peptides are then screened for activity without removing them from the pins.
Houghton, 1985, Proc. Natl. Acad. Sci. USA, 82, 5131, and U.S. Pat. No. 4,631,211, utilizes individual polyethylene bags (xe2x80x9ctea bagsxe2x80x9d) containing C-terminal amino acids bound to a solid support. These are mixed and coupled with the requisite amino acids using solid phase synthesis techniques. The peptides produced are then recovered and tested individually.
Fodor et al., 1991, Science, 251, 767, described light-directed, spatially addressable parallel-peptide synthesis on a silicon wafer to generate large arrays of addressable peptides that can be directly tested for binding to biological targets. These workers have also developed recombinant DNA/genetic engineering methods for expressing huge peptide libraries on the surface of phages. Cwirla et al., 1990, Proc. Natl. Acad. Sci. USA, 87, 6378.
In another combinatorial approach, V. D. Huebner and D. V. Santi (U.S. Pat. No. 5,182,366) utilized functionalized polystyrene beads divided into portions each of which was acylated with a desired amino acid; the bead portions were mixed together, then divided into portions each of which was re-subjected to acylation with a second desirable amino acid producing dipeptides, using the techniques of solid phase peptide synthesis. By using this synthetic scheme, exponentially increasing numbers of peptides were produced in uniform amounts which were then separately screened for a biological activity of interest. Another method of producing libraries of organic compounds based on dipeptides, hydantoins and benzodiazepines using a polystyrene based solid support is described by DeWitt et al. 1993, Proc. Natl. Acad. Sci. USA, 90, 6909.
Bunin et al., 1992, J. Am. Chem. Soc., 114, 10997, describe a method for the combinatorial synthesis of large libraries of peptides. According to Bunin, 2-amino benzophenones are attached to a polystyrene solid support and converted into various 1,4 benzodiazepine derivatives, which can then be screened for specific receptor or enzyme activity.
Zuckerman et al., 1992, Int. J. Peptide Protein Res. 91, 1 and 1993, Structural Biology, 3, 580, also have developed similar methods for the synthesis of peptide libraries and applied these methods to the automation of a modular synthetic chemistry for the production of libraries of, for example, N-alkyl glycine peptide derivatives, called 44 peptoidsxe2x80x9d, which are screened for activity against a variety of biochemical targets. See also, Symon et al., 1992, Proc. Natl. Acad. Sci. USA, 89, 9367. Encoded combinatorial chemical syntheses have been described recently. Brenner et al., 1992, Proc. Natl. Acad. Sci. USA, 89, 5381.
The focus of these structural diversity activities on peptide synthesis chemistry is a direct result of the fact that the ability to generate structural diversity requires, as its starting point, the access to practical stepwise sequential synthesis chemistries that allow the incorporation of varied structural elements with orthogonal reactivities. To date, these have only been worked out for the Merrifield synthesis of peptides and the Carruthers synthesis of oligonucleotides. Thus, there remains a need for an improved method for the structure-directed generation and screening of organic compounds to determine which may be suitable in a particular application.
One group of compounds that allow the incorporation of varied structural elements are xcex1-keto acids, esters, and amides. The xcex1-keto esters and amides are known to be potent inhibitors of proteolytic enzymes, such as serine and cysteine protease, as well as showing inhibition of leukotriene A4 hydrolase and chymase, Wasserman et al., (Cyanomethylene)phosphoranes as Novel Carbonyl 1,1-Dipole Synthons: An Efficient Synthesis of xcex1-Keto Acids, Esters, and Amides, J. Org. Chem 1994, 59, 4364-4366 (xe2x80x9cWasserman 1994xe2x80x9d). The biological activity of these compounds is believed to be the result of the presence of the electron-deficient xcex1-keto group, which is similar in reactivity to the carbonyl group of xcex1-fluorinated inhibitors and the xcex1- and xcex2-carbonyl groups in vicinyl tricarbonyls.
Vicinyl tricarbonyls include xcex1,xcex2-diketo amides, which are more reactive than xcex1-keto amides, and are found in a number of biologically active, naturally occurring peptide analogues, such as the immunosuppressants FK-506, rapamycin, and the elastase inhibitors YM-47141 and YM-47142. Because of their biological activity and potential as immunosuppressants and inhibitors, a method for the synthesis of xcex1,xcex2-diketo acids, esters and amides is highly desirable.
Wasserman 1994 teaches a method for the synthesis of xcex1-keto acids, esters, and amides. A ylide, (cyanomethylene)triphenylphosphorane, undergoes a coupling reaction with a carboxylic acid to form a cyano keto phosphorane. The cyano keto phosphorane is readily oxidized to form a highly electrophilic vicinal diketo nitrile that can be trapped at low temperature by reaction with nucleophiles to form a transient cyanohydrin intermediate. The cyanohydrin readily undergoes elimination of hydrogen cyanide to form an xcex1-keto acid, ester or amide. Whether an xcex1-keto acid ester or amide is formed is determined by the choice of nucleophile. Structural diversity is obtained by the appropriate choice of the carboxylic acid and nucleophile.
Wasserman et al., 1993 Synthesis and Evaluation of Peptidyl Vicinal Tricarbonyl Monohydrates as Inhibitors of Hydrolytic Enzymes, J. Org. Chem., 58, 4785-4787 (xe2x80x9cWasserman 1993xe2x80x9d), disclose peptidyl vicinal tricarbonyls, prepared from N-protected di- and tripeptides by reaction of carboxylic acid residues with ylides, followed by oxidation. The peptide vicinal tricarbonyls have been shown to be potent inhibitors of serine protease.
However, there is no teaching in the prior art of a method for the synthesis of xcex1,xcex2-diketo acids, esters, or amides from xcex1-keto acids. The present invention provides such a method.
The present invention relates to a method for the synthesis of vicinyl tricarbonyl compounds, and, in particular, xcex1,xcex2-diketo acids, esters, and amides, i.e., vicinyl tricarbonyl compounds of formula 
where R is a structural diversity element selected from the group consisting of alkyl, cycloalkyl, aryl, heteroaryl, peptidyl, heteroatom-substituted alkyl, cycloalkyl, alcohols, and amines; and Nu is a structural diversity element derived from a nucleophile, NuH, selected from the group consisting of amines, amino acids, peptides, water, hydrogen sulfide, alcohols, and thiols.
The method of the invention comprises reacting an xcex1-keto acid with a ylide, preferably a (cyanomethylene)phosphorane, such as a triphenylphosphorane, under sufficient reaction conditions (i.e., pressure, temperature, and time, and preferably in the presence of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride) to form a cyano diketo phosphorane, oxidizing the cyano diketo phosphorane, preferably with ozone, forming a cyano tricarbonyl and reacting the cyano tricarbonyl with a nucleophile to form the xcex1,xcex2-diketo acid, ester, or amide. Preferably, the nucleophile is an amine, amino acid, peptide, hydrazine, water, hydrogen sulfide, alcohol, or thiol. Typically, the cyano diketo phosphorane is oxidized with ozone in CH2Cl2 at a temperature of less than about xe2x88x9260xc2x0 C., preferably about xe2x88x9278xc2x0 C., and the cyano tricarbonyl is reacted with the nucleophile for about 2 to about 60 minutes at a first temperature of less than about xe2x88x9260xc2x0 C., preferably about xe2x88x9278xc2x0 C., followed by from about 30 minutes to about 4 hours at a second temperature of less than about 5xc2x0 C., preferably about 0xc2x0 C.
The xcex1-keto acid is typically formed by reacting a ylide, again, preferably a (cyanomethylene)phosphorane, such as a triphenylphosphorane, under sufficient reaction conditions with a carboxylic acid or acid chloride to form the corresponding cyano keto phosphorane, oxidizing the cyano keto phosphorane, again, preferably with ozone, forming a vicinyl diketo nitrile; and trapping the vicinyl diketo nitrile with water, forming an xcex1-keto acid. Preferably, the carboxylic acid contains a structural diversity element of an alkyl, cycloalkyl, aryl, heteroaryl, peptidyl, heteroatom-substituted alkyl, cycloalkyl, or amine group. The carboxylic acid can be an amino acid.
The nucleophile used to form xcex1,xcex2-diketo acid, ester, or amide of the invention may be an alcohol comprising from 1 to about 10 carbon atoms, an amino acid selected from the group consisting of tryptophan, arginine, histidine, glutamic acid, glutamine, aspartic acid, leucine, threonine, proline, alanine, tyrosine, carbamido cysteine, phenylalanine, methionine, lysine, asparagine, isoleucine, cysteine, valine, serine, and glycine, an amine comprising at least one alkyl or cycloalkyl group comprising 1 to about 10 carbon atoms, or a heterocyclic ring compound comprising at least one nitrogen atom and from 3 to about 10 carbon atoms in the ring.
Arrays may be formed by making compounds according to the invention, and spatially arranging a plurality of such compounds to form the array. Typically, an mxc3x97p array of q xcex1,xcex2-diketo molecules or an mxc3x97p array of compartments is formed, where each compartment contains an xcex1,xcex2-diketo compound, and where m and p are integers representing the number of rows and columns in the array, and q is an integer in the range of from 1 to the product of m multiplied by p, and represents the number of different compounds in the array. The product of m multiplied by p is typically at least about 25, but is preferably at least about 3,000, and may exceed 10,000. The array may then be used in a combinatorial library of compounds, comprising r different xcex1,xcex2-diketo compounds, wherein r is an integer greater than 1, typically greater than 25, and may exceed 100,000.
The invention also relates to a combinatorial library of compounds, comprising r different compounds of formula 
wherein R is a structural diversity element selected from the group consisting of alkyl, cycloalkyl, aryl, heteroaryl, peptidyl, heteroatom-substituted alkyl, heteroatom-substituted cycloalkyl, and amines; Nu is a structural diversity element derived from a nucleophile, NuH, by removal of a hydrogen atom, wherein NuH is selected from the group consisting of amines, amino acids, peptide, water, hydrogen sulfide, alcohols, and thiols; and r is an integer of 2 to 96.
Preferably, r is an integer greater than 25, and at least one of R or Nu is derived from an amino acid by removal of a hydrogen atom, where Nu may be an amino acid selected from the group consisting of tryptophan, arginine, histidine, glutamic acid, glutamine, aspartic acid, leucine, threonine, proline, alanine, tyrosine, carbamido cysteine, phenylalanine, methionine, lysine, asparagine, isoleucine, cysteine, valine, serine, and glycine, and R may be an amino acid selected from the group consisting of arginine, glutamic acid, glutamine, aspartic acid, leucine, threonine, proline, alanine, tyrosine, phenylalanine, lysine, asparagine, isoleucine, valine, serine, and glycine.
In an alternate embodiment, the invention relates to a method for making an array of compounds. The method comprises preparing compounds according to the following method: 
wherein R1 is a structural diversity element selected from the group consisting of alkyl, cycloalkyl, aryl, heteroaryl, peptidyl, heteroatom-substituted alkyl, cycloalkyl, and amines; and Nu is a structural diversity element derived from a nucleophile, NuH, by removal of a hydrogen atom, wherein NuH is selected from the group consisting of amines, amino acids, peptide, water, hydrogen sulfide, alcohols, and thiols; and spatially arranging a plurality of such compounds to form the array.
Preferably, R1 is derived from an amino acid by removal of a hydrogen atom, where the amino acid is selected from the group consisting of arginine, glutamic acid, glutamine, aspartic acid, leucine, threonine, proline, alanine, tyrosine, phenylalanine, lysine, asparagine, isoleucine, valine, serine, and glycine, the oxidizing step is carried out with ozone, the nucleophile NuH is an alcohol, a thiol, an amine, an amino acid, a peptide, or hydrazine.
The invention also relates to a method for the synthesis of a triacyl diamide. The method of the invention comprises reacting a secondary amine with an alkyl oxalyl chloride to form an oxalyl ester, hydrolyzing the oxalyl ester in LiOH/THF to a carboxylic acid, reacting the carboxylic acid with a (cyanomethylene)phosphorane to form a cyano diketo phosphorane, oxidizing the cyano diketo phosphorane to produce a triacyl nitrile intermediate that is susceptible to attack by a nucleophile, and reacting the triacyl nitrile intermediate with a second amine, preferably a secondary amine to form the triacyl diamide.
The present invention relates to a method for the synthesis of vicinyl tricarbonyl compounds, and, in particular, xcex1,xcex2-diketo acids, esters, and amides, and to the formation of combinatorial libraries containing xcex1,xcex2-diketo compounds.
The vicinyl tricarbonyl compounds of the invention are synthesized from xcex1-keto acids produced in accordance with the method disclosed by Wasserman 1994 and in copending U.S. patent application Ser. No. 08/503,070, the content of which is incorporated in its entirety by reference. The xcex1-keto acid is subjected to a coupling reaction with a (cyanomethylene)phosphorane, such as a triphenylphosphorane, in the presence of 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (xe2x80x9cEDCIxe2x80x9d) to form a cyano diketo phosphorane. The cyano diketo phosphorane is then oxidized to form a highly electrophilic vicinyl triketo nitrile that is trapped with a nucleophile to form a cyanohydrin intermediate. The cyanohydrin then undergoes facile elimination of hydrogen cyanide to form an xcex1,xcex2-diketo acid, ester, or amide, depending on the nucleophile used.
More specifically, a (cyanomethylene)phosphorane, 1, a ylide, is made to undergo a coupling reaction with a carboxylic acid or acid chloride of formula 2 to form a cyano keto phosphorane of formula 3, as shown below. 
The cyano keto phosphorane, 3, is then oxidized, preferably with ozone (O3) to form a vicinyl diketo nitrile, 4. Upon nucleophilic attack by water, the vicinyl diketo nitrile, 4, forms a cyanohydrin intermediate of formula 5. The cyanohydrin intermediate 5 readily eliminates hydrogen cyanide to form an xcex1-keto acid, 6. 
The xcex1-keto acid, 6, is then converted to an xcex1,xcex2-diketo acid, ester, or amide by first making the xcex1-keto acid, 6, undergo a second coupling reaction with a (cyanomethylene)phosphorane to form a cyano diketo phosphorane, 7, which is oxidized, preferably by ozone, to form a cyano tricarbonyl, 8. On reaction of the cyano tricarbonyl, 8, with a nucleophile of formula NuH, an intermediate is formed that readily eliminates hydrogen cyanide to form an xcex1,xcex2-diketo acid, ester, or amide. 
Whether the resulting xcex1,xcex2-diketo compound is an acid, ester, or amine depends on the nucleophile used. For example, where the nucleophile, NuH, is water, an xcex1,xcex2-diketo acid of formula I is formed. 
Similarly, where nucleophile NuH is an alcohol of formula R2OH, an xcex1,xcex2-diketo ester of formula II is formed, 
and where nucleophile NuH is an amine of formula HNR3R4, an xcex1,xcex2-diketo amide of formula III is formed. 
Groups R1, R2, R3, R4, and Nu are structural diversity elements that are specifically chosen for the properties and structure that they provide to the resulting xcex1,xcex2-diketo compound. The R groups, R1, R2, R3, and R4, may be the same or different, and include, but are not limited to alkyl, cycloalkyl, substituted and unsubstituted aryl, heteroaryl, heteroatom-substituted alkyl and cycloalkyl, amidyl, peptidyl, and alkoxyl. Useful Nu groups are derived from nucleophiles of formula NuH that include, but are not limited to amines (including hydrazines and primary and secondary amines), amino acids, peptides, water, hydrogen sulfide, alcohols, thiols, and carbon-centered nucleophiles, such as indoles, enamines, enols, enolates, silyl enol ethers, ethers, cuprates and other metallated species. Useful peptides include peptides containing from 2 to about 50, preferably 2 to about 25 amino acids. Useful amino acids include, but are not limited to, tryptophan, arginine, histidine, glutamic acid, glutamine, aspartic acid, leucine, threonine, proline, alanine, tyrosine, carbamido cysteine, phenylalanine, methionine, lysine, asparagine, isoleucine, cysteine, valine, serine, and glycine.
As can be seen from the reaction scheme set forth above, in one embodiment, structural diversity element R1 is determined by the selection of the carbonyl compound of formula 2 that undergoes the coupling reaction with the ylide, 1, to form cyano keto phosphorane, 3. Similarly, structural diversity elements R2, R3, R4, and Nu are determined by the choice of the nucleophile selected to trap the cyano tricarbonyl, 8, which then eliminates hydrogen cyanide to form the desired xcex1,xcex2-diketo compound.
In an alternate embodiment, tricarbonyl derivatives can be formed by converting a secondary amine to an oxalyl ester, which is then hydrolyzed in LiOH/THF to the corresponding carboxylic acid. The carboxylic acid is then coupled with a (cyanomethylene)phosphorane for the oxidative incorporation of a third carbonyl group. The triacyl nitrile intermediate is then made to react with a nucleophile, as in the scheme described above, to form the desired xcex1-xcex2-diketo compound.
This process may be used to form novel triacyl diamides. A secondary amine, 10, is made to react with an ester-acid chloride of formula 11 to form an oxalyl ester, 12, which is hydrolyzed in LiOH/THF to the corresponding carboxylic acid, 13. The xcex1-keto ester may be synthesized using the method described above for the synthesis of xcex1-keto acids by using an alcohol of formula ROH as the nucleophile. 
The carboxylic acid, 13, is then made to undergo the coupling reaction with a (cyanomethylene)phosphorane in EDCI, as described above, to form a cyano diketo phosphorane of formula 14, as shown below. 
Oxidation of the cyano diketo phosphorane, 14, produces a triacyl nitrile intermediate, 15, that is susceptible to attack by a nucleophile. 
Where the nucleophile is a secondary amine, the resulting xcex1-xcex2-diketo compound is a triacyl diamide, 16. It has been discovered that the reaction of the triacyl nitrile, 15, with the nucleophile occurs at the xcex1-carbonyl, leading to the formation of the tricarbonyl derivative, 16. 
Prior to the present invention, it was believed that attack by the nucleophile could occur at the xcex2-carbonyl, leading to the elimination of carbon monoxide and hydrogen cyanide, and the formation of an xcex1-keto compound. 
Although this result is obtained in certain cases that have geometric features that favor attack by the nucleophile at the xcex2-carbonyl, the reaction typically occurs exclusively at the xcex1-carbonyl. This is significant, because it is believed that the xcex2-carbonyl should be more electrophilic than the xcex1-carbonyl. Without being bound by theory, it is believed that the xcex2-attack may be reversible, while the attack at the xcex1-carbonyl results in the irreversible elimination of cyanide. Similar results are obtained with nucleophiles other than amines, e.g., alcohols and water, resulting in the formation of xcex1,xcex2-diketo esters and acids, as well as the triacyl amides formed where the nucleophile is an amide, as shown above.
The ability of these various reactions to be carried out in a stepwise sequential process using modules chosen in a structure-directed manner allows the production of structurally directed thematic diversity libraries, having, structural elements systematically varied around a basic motif.
Combinatorial libraries of xcex1,xcex2-diketo compounds may be synthesized by the modular development of xcex1,xcex2-diketo acids, esters, and amides that have selected properties. Once the library has been formed, the xcex1,xcex2-diketo compounds that make up the library may be screened to determine which compounds best meet the requirements of a particular application.
The present invention may be used to generate a number of different molecules for screening purposes by first forming an xcex1,xcex2-diketo compound as a base module having at least two structural diversity elements attached. By fixing one of the positions and structures of the structural diversity elements, and by varying at least one of the others, an array of different xcex1,xcex2-diketo molecules is easily generated. These molecules can then be screened to determine which are suitable for a particular application or target use. Once a suitable xcex1,xcex2-diketo compound is identified, it can be selected for generating a further array of molecules. This is done by modifying the particular structural diversity elements that are found to be suitable, or by combining the chosen structural diversity element with an expanded or different set of second compounds or elements. This process can be repeated as often as necessary to develop the optimum compound for the particular use.
The particular xcex1,xcex2-diketo base module chosen for use in accordance with the present invention is not critical, and can be any one of a wide variety of structures. Knowledge of the base modules can be represented in the form of combinatorial libraries.
From the foregoing, it is seen that various arrays of xcex1,xcex2-diketo molecules can be prepared. These arrays can be generated in any size desired to facilitate the screening of a large number of molecules at one time, and are preferably spatially arranged. For example, standard arrays having 96 compartments in an 8xc3x9712 array can be used, where any number of compartments in the array contain different molecules, while the other can contain controls or duplicate samples. Preferably, in an 8xc3x9712 array, 16 of the compartments contain controls and 80 compartments contain different samples. After an initial screening identifies xcex1,xcex2-diketo molecules having certain beneficial or desirable properties, a second tray containing, e.g., 20 samples of each of 4 different xcex1,xcex2-diketo molecules, again with 16 control samples, can be used to confirm the original results. The samples can be placed in columns of the same material, or a completely random array can be generated to have a completely blind analysis.
In view of these variations, one of ordinary skill in the art would understand that any mxc3x97p array of molecules can be generated, where m and p are integers, m being greater than zero, and p being greater than 1. There is no upper limit to m and p other than the capabilities of the testing or screening equipment. As noted above, an 8xc3x9712 array is typical, but q compounds can be tested from arrays where m or p is as high as 25 or more; q being an integer from 1 up to the total of m times p, and typically being between 2 and 96, although significantly larger arrays are contemplated. At this time, it is specifically preferred that m and p be integers of between 3 and 15, and that a few control molecules be included so that q is less than the product of m and p. However, this invention contemplates that use of any integer for m or p, with each integer or combination of mxc3x97p integers relied upon as representing a useful embodiment. Thus, q may be an integer equal to 1 up to the product of m multiplied by p.
As noted above, the xcex1,xcex2-diketo molecules used in the array would be generated from one or more of the xcex1,xcex2-diketo base molecules described herein. In this manner, combinatorial libraries of r different compounds, where r is any integer greater than 1, can be made. Typically, r will be greater than 5, preferably at least 25. As noted, r can be as high as 80 or 96 using available trays, or can even be any higher number using multiple or specifically designed trays. Although for convenience, linear arrays are described, the specific arrangement of the molecules and tray compartments can be circular, staggered or in any other configuration which can be analyzed by the testing or screening device used.
In one embodiment of the present invention at least two of the structural diversity elements, R1, R2, R3, R4, or Nu, are present, and one or more are reactive groups that are capable of further reactions to produce a base module. For example, the present invention is directed to structural diversity groups that may themselves be capable of further reaction to form base modules as described herein.
The determination of compounds that meet the requirements of a particular application involves a process, which comprises: a) the synthesis of an array of different xcex1,xcex2-diketo acids, esters, and amides, each containing substituent groups, selected to provide structural diversity to the elements of the array, and/or the reaction of xcex1,xcex2-diketo compounds with other reactive chemical species to alter or exchange substituent groups on the xcex1,xcex2-diketo compounds to produce an array, each element of which is a compound consisting of molecules having a selected set of substituent groups, such that each element of the array is a compound having selected properties that are determined by the selected substituent groups on the xcex1,xcex2-diketo acid, ester or amide molecules that make up the compound; and b) the screening of at least some of the elements of the array to determine which compounds in the array have properties that meet the requirements of a particular application.
The difference between the elements in the array is determined by the choice of structural diversity elements, i.e., the substituent group R1, introduced into the xcex1,xcex2-diketo compound by carboxylic acid or acid chloride 2, and the substituent group Nu, introduced into the xcex1,xcex2-diketo compound by nucleophile NuH. As will be readily appreciated by one of ordinary skill in the art, the presence and the structure of structural diversity elements R2, R3, and R4 are determined by the choice of nucleophile NuH. By the appropriate choice of R1 and Nu, the resulting xcex1,xcex2-diketo compounds can be designed to have specific, pre-selected properties. For example, as described above, the choice of nucleophile determines whether the xcex1,xcex2-diketo compound will be an acid, an ester or an amide, where water provides an acid of formula I, an alcohol provides an ester of formula II, and an amine provides an amide of formula III. By varying both R1 and Nu, an array of xcex1,xcex2-diketo compounds can be formed. Typically, each row of the array would have the same R1 group with differing Nu groups, and each column would have the same Nu group with differing R1 groups.
As used herein, the phrase alkyl means any branched or straight chain, substituted or unsubstituted acyclic carbon-containing compounds, including alkanes, alkenes and alkynes, typically containing up to about 30 carbon atoms. Examples of alkyl groups include lower alkyl, for example, methyl, ethyl, n-propyl, iso-propyl, n-butyl, iso-butyl or tert-butyl; upper alkyl, for example, octyl, nonyl, decyl, and the like; and lower alkylene, for example, ethylene, propylene, propyldiene, butylene, butyldiene, pentene, hexene, heptene, octene, norbornene, nonene, decene and the like. The ordinary skilled artisan is familiar with numerous linear and branched alkyl groups, which are within the scope of the present invention.
In addition, such alkyl groups may also contain various substituents in which one or more hydrogen atoms has been replaced by a functional group. Functional groups include, but are not limited to hydroxyl, amino, carboxyl, sulfonic amide, ester, ether, phosphates, thiol, nitro, silane, and halogen, i.e., fluorine, chlorine, bromine and iodine, to mention but a few. Substituted alkyl groups include, but are not limited to, alkoxy, e.g., methoxy, ethoxy, butoxy, and pentoxy, amino, e.g., dimethylamino, diethylamino, cyclopentylamino, benzylmethylamino, and dibenzylamino, amido, e.g., formamido, acetamido, and butanamido.
As used herein, cycloakyl means a substituted or unsubstituted cyclic carbon-containing compound, including, but not limited to, cyclopentyl, cyclohexyl, cycloheptyl, adamantyl, and the like. Such cyclic groups may also contain various substituents in which one or more hydrogen atoms has been replaced by a functional group. Such functional groups include those described above, and lower alkyl groups having from 1-28 carbon atoms. The cyclic groups of the invention may further comprise at least one heteroatom, typically a nitrogen forming a cyclic secondary amine, typically containing up to about 10 carbon atoms. Heterocyclic ring compounds containing more than one heteroatom are also useful in the invention.
As used herein, aryl groups means a substituted or unsubstituted hydrocarbon ring, bearing, a system of conjugated double bonds, comprising 4n+2xcfx80-bond electrons, where n is an integer greater than or equal to 0. Examples of aryl groups include, but are not limited to, phenyl, naphthyl, anisyl, toluyl, xylenyl and the like. According to the present invention, aryl also includes aryloxy, aralkyl, aralkyloxy and heteroaryl groups, e.g, pyrimidine, morpholine, piperazine, piperidine, benzoic acid, toluene or thiophene and the like. These aryl groups may be substituted with any number of a variety of functional groups. In addition to the functional groups described above in connection with substituted alkyl groups and carbocyclic groups, functional groups on the aryl groups can be nitro groups.
As mentioned above, these structural moieties can also be any combination of alkyl, carbocyclic or, aryl, groups, for example, 1-cyclohexylpropyl, benzylcyclohexylmethyl, 2-cyclohexylpropyl, 2,2-methylcyclohexylpropyl, 2,2-methylphenylpropyl, 2,2-methylphenylbutyl, and the like.